Maritime Threat Assessment · 10 September 2026

Red Sea Threat Now Turns on Geography as Much as Affiliation

September 2026 reassessment. The Houthi capture of Mokha and continued strikes on Saudi-affiliated shipping and energy infrastructure have deepened an affiliation-led threat and made southern Red Sea geography a central variable. Regional threat does not equal individual vessel risk.

Trident Risk & AdvisoryRed Sea / Gulf of AdenMaritime SecurityGeopolitical Analysis

Key judgement

On the evidence retained at this cut-off, Houthi targeting remains affiliation-led: the strikes centre on Saudi-flagged, Saudi-operated and Saudi-associated tonnage and on the west-coast energy corridor, alongside the standing Israel, US and UK target set. The material new variable is geography. The capture of Mokha compresses reaction time in the southern Red Sea without conferring control of Bab el-Mandeb, where the decisive terrain (Perim Island and the Dhubab coast) is contested and unconfirmed. A regional rating cannot be transited: exposure resolves only vessel by vessel.

01Executive assessment

The 13 August edition captured a Houthi naval blockade of Saudi Arabia (declared 20 July) and an affiliation-led target set built around Saudi-flagged and Saudi-operated tonnage. Since then the picture has not reversed; it has deepened and acquired a geographic dimension. This assessment carries an intelligence cut-off of 10 September 2026 (1200Z).

Through late August the campaign continued to strike Saudi-affiliated shipping and energy infrastructure, including the Bahri tanker Amzan near Yanbu on 24 August, confirming reach into the northern approaches. In early September Houthi forces intensified strikes on southern Saudi energy sites, and on 10 September they captured the Red Sea port city of Mokha, about 75 km north of Bab el-Mandeb, from pro-government forces.

Exposure now turns on two axes rather than one: which vessels the campaign is trying to hit (affiliation), and where a vessel is most exposed (geography). A regional rating cannot be transited; exposure resolves only vessel by vessel.

02What has changed since 13 August

Three developments have genuinely changed the picture since the 13 August edition, each occurring after that date. The Bahri (Saudi) tanker Amzan was struck by a ballistic missile about 63 nautical miles west of Yanbu on 24 August, confirmed by UKMTO and the owner, with the crew reported safe: evidence that the campaign's reach extends to the northern approaches and that the focus remains on Saudi-affiliated tonnage.

In early September, Houthi forces intensified drone and missile strikes on southern Saudi energy sites (Jazan, Abha and Najran), with reciprocal Saudi airstrikes. On 10 September they captured Mokha (and Zuqar Island) from pro-government forces, extending Houthi coastal reach about 75 km north of Bab el-Mandeb.

Two standing features carry forward and continue to matter for planning: the Joint War Committee's widened Red Sea Listed Area (JWLA-034, 29 July, notification line to about 25.5°N), and persistent GNSS jamming and spoofing across the corridor with attribution unresolved.

03Targeting: affiliation remains the discriminator

On the observed record, Saudi flag, ownership, operation and association remain the clearest discriminators of Houthi targeting, alongside the standing Israel, US and UK target set; voyages to or near the Yanbu and Jizan energy corridor carry both targeting and collateral-infrastructure exposure. Vessel type is not an observed discriminator. This is why a defensible transit view begins with a vessel affiliation assessment rather than ship class.

The retained 2026 record is consistent with an affiliation-led model rather than a broadening of the target set. Mistaken identity remains a residual possibility, on the 2024 precedent of vessels struck on the basis of out-of-date associations, so a clean current profile should be tested for historic or apparent links rather than assumed safe. Commercial relationships, such as charterer or operator, may contribute to how a vessel is perceived where a counterparty is associated with the target set, though the observed record is not yet sufficient to treat that as an independently evidenced discriminator.

04Geography: a gradient, and why Mokha did not close the strait

The word 'Red Sea' is too coarse for a transit decision. Kinetic exposure rises moving south: lowest in the northern Red Sea, rising through the central Red Sea (within reach of Hodeidah-based weapons), and peaking across the southern Red Sea and Bab el-Mandeb. The strait is about 18 nautical miles at its narrowest, split by Perim Island into two roughly two-mile channels.

The capture of Mokha must be read precisely, because it is easy to overstate. A southern-coast foothold plausibly improves coastal observation and shortens launch geography for missiles, drones and explosive small craft, compressing warning time. It does not confer control of Bab el-Mandeb: Mokha lies about 75 km north of the strait and its guns cannot range the waterway. The decisive terrain is Perim Island, which splits the strait, and the Dhubab coast; reports of a Houthi landing on Perim are contested and unconfirmed at the cut-off, and this is the single most important indicator to watch.

05The mine threat: capability and history, not a confirmed campaign

Because Houthi naval-mine capability plainly exists, it is tempting to assume an active mine threat. The evidence does not support that. Capability is demonstrated (the coalition has reported destroying around 86 Houthi-planted mines) and there is historic harm (drifting mines between Midi and Jizan sank an Egyptian fishing vessel in 2020, killing three).

But through the intelligence cut-off there is no confirmed mine strike on a commercial vessel and no evidenced new mass mine-laying against commercial traffic, only a single-source, partially inaccessible August report of a mine removed from the Bab el-Mandeb lane. The honest current assessment is that the mine threat is latent: a credible capability and pattern, not an active deployment.

06The transit environment: reporting, not guaranteed protection

The naval and advisory architecture is real but frequently misread as a guarantee. UKMTO is the primary voluntary reporting point; EUNAVFOR ASPIDES provides monitoring and, on request, accompaniment and close protection, but it has cautioned that available assets have not increased, so protection can involve waiting time and should not be assumed in planning. ASPIDES is defensive; it does not strike shore targets. It has advised Israeli-, US- and Saudi-linked vessels to avoid transiting the corridor.

For owners and charterers, the operative posture is a ship- and voyage-specific plan: register and report with UKMTO, treat GNSS as potentially degraded, and confirm war-risk arrangements against the widened Listed Area before committing a transit. These are the questions answered by high-risk area transit planning and a voyage vulnerability assessment.

07From regional threat to vessel decision

A regional threat picture is the backdrop to a decision, not the decision. Two ships transiting the same waters on the same day can carry very different risk. Converting this assessment into a voyage decision means locating the specific vessel within it: a vessel affiliation assessment to test how it is likely to be perceived, a voyage vulnerability assessment to place the passage within the geographic gradient, and high-risk area transit planning with bridge response support for the transit itself.

The full assessment, including the rebuilt targeting model, the geographic exposure model, the threat-vector and mine-threat analysis, the attack-pattern timeline and the sourcing register, is available to download. Where a matter turns on whether a particular vessel faced real danger, the same method supports war-risk expert evidence in charterparty disputes.

Full report

The complete assessment, including its detailed findings, is available to download.

Need advice on this issue?

Speak to an analyst about how this applies to a specific vessel, voyage, itinerary or dispute.

Speak to an analyst